The proposed research is designed to investigate factors determining the degree of overlap of conceptual representations of the two languages of bilingual speakers. The proposal compares highly proficient and less proficient bilingual speakers, and aims to expand on previous research in three ways: 1) by examining a wider range of word classes; 2) by exploring complete and partial translation equivalency; 3) by addressing the possibility of conceptual restructuring in the native language (L1) as a consequence of exposure to a second language (L2). Exps. 1&2 examine these issues in out-of-context language processing tasks, including translation, lexical decision and naming, and will lay the ground for later experiments. Exps. 3&4 will examine the consequences of conceptual overlap in contextual processing, using behavioral reaction time measures. Exp. 5 will use measures of eye-movements during reading, and will provide converging evidence or previous findings and a better understanding of their time course. The guiding hypothesis is that L1 and L2 conceptual representations are not entirely shared, and that representations constructed during L1 acquisition might be modulated by the conceptual structure of L2, especially for highly proficient bilinguals.